The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> help needed on wideband matching
https://designers-guide.org/forum/YaBB.pl?num=1207009725

Message started by chase.ng on Mar 31^st, 2008, 5:28pm

Title: help needed on wideband matching
Post by chase.ng on Mar 31^st, 2008, 5:28pm

Hi all,

I am not very familiar with impedance matching. I have a circuit with its input impedance that is a low pass equivalent to 1k // 2pF. It is connected to a 50 ohm source via a 50 ohm tline. I would like to match the 1k // 2pF input to 50 ohm so that the source always sees a 50 ohm load. The problem is I need the match to work from 1 GHz to 2 GHz. Any idea? Please assist.

thanks and regards,
chase

Title: Re: help needed on wideband matching
Post by pancho_hideboo on May 4^th, 2008, 11:53am

chase.ng wrote on Mar 31^st, 2008, 5:28pm:

I have a circuit with its input impedance that is a low pass equivalent to 1k // 2pF.

I don't know what you mean by "a low pass equivalent to 1k // 2pF".
But if it means simple parallel RC impedance, simple resistive matching is most reasonable, I think.

Or do you mean an equivalent low pass representation of bandpass impedance ?

Title: Re: help needed on wideband matching
Post by pancho_hideboo on May 5^th, 2008, 12:40pm

From Bode-Fano criteria for Parallel RC load matched with reactive networks,
the best reflection coefficient averaged over 1GHz ~ 2GHz is like following.

Averaged_Gamma >= exp{-1/(2*BW*R*C)}, Here BW=2GHz-1GHz=1GHz, R=1kOhm, C=2pF.

Then Averaged_Gamma_dB >= -2.17dB
dB(S(5,5)) can almost achieve this theoretical limit.

Title: Re: help needed on wideband matching
Post by chase.ng on May 10^th, 2008, 7:36pm

Hi,

Thanks for your help. I am still quite new to the RF matching so I would need some time to digest the matching network you provided. Yes. I mean it is a simple parallel RC.

But a few question pop up in my mind that I hope you can shed some light. Isn't that the best matching condition will yield very small S11 number? -2.17 still seems too high. Does that means i have no choice but to use resistive matching?

Many thanks... :D

Regards,
Chase.

Title: Re: help needed on wideband matching
Post by pancho_hideboo on May 14^th, 2008, 7:40am

chase.ng wrote on May 10^th, 2008, 7:36pm:

Isn't that the best matching condition will yield very small S11 number? -2.17 still seems too high.

It depends on your applications.
If you have importance in reflection, this return loss is too large.
But if you have importance in transmitted power, this return loss could be acceptable.

chase.ng wrote on May 10^th, 2008, 7:36pm:

Does that means i have no choice but to use resistive matching?

This also depends on your applications.

If you invoke hybrid matching of resistive and reactive, you can achieve better S11.

What application do you intend to design ?
What signal scheme do you apply ?
Is there frequency channel control signal. e.g. for PLL ?
Is there any possibility to change matching networks using such control signal ?

Varactor, Pin-diode, etc...